A major challenge in the management of lung disease is the development of multi-variate non-invasive approaches to predict clinical outcome and response to therapy. One important approach, already widely used in pulmonary medicine, is volumetric computed tomography (CT);however, utilization of the CT data is suboptimal. Clinical evaluation of CT images is mostly subjective and qualitative, resulting in spatial, temporal, and inter-observer variability. The unique features of the lung, i.e., its non-solid, elastic architecture, render the organ highly sensitive to volume change and to asymmetric distortion due to regional disease involvement;architectural distortion can make it difficult to match the same anatomical regions on successive scans, particularly if lung inflation state has changed. There is a need for objective analytical strategies that fully exploit the information content of the CT dataset to permit earlier detection and improve diagnostic yield in the assessment of lung pathology. Quantitative analysis has not been widely adopted by pulmonologists and radiologists, partly because of the perceived "cumbersome" nature of analysis, and partly because of the lack of concrete examples that demonstrate how quantitative analysis facilitates detection of specific pathology. We have developed and tested a relatively simple PC-based semi-automated method for separating lung lobes and mapping regional CT attenuation within and among lobes. From voxelwise attenuation values the distributions of absolute and fractional air and tissue volumes are derived, and compared within and among lobes along standardized 3D coordinate axes. We will test the hypothesis that quantification of the magnitude, spatial distribution and heterogeneity of regional CT-derived parameters provides objective biomarkers that reflect clinical and functional abnormalities in diffuse lung disease. We will examine two Specific Aims: 1) quantify, map and compare CT-derived parameters within and among lobes in normal subjects and in patients with different degrees of diffuse parenchymal pathology: emphysema and interstitial lung disease, b) correlate intra- and inter-lobar distributions of CT-derived parameters with indices of clinical and functional impairment. These studies will establish or refute the practical utility of quantitative analysis in evaluating clinical chest CT in relation to lung pathology;they provide the essential data that may lead to future larger-scale prospective studies to test whether this approach accurately tracks or predicts longitudinal disease progression or response to treatment. The extensive CT database maintained by LTRC provides a unique and valuable resource for addressing these issues. Given the widespread use (and possible overuse) of CT in pulmonary medicine and the small but real risk of harm associated with repeated radiation exposure, it is imperative to develop and adopt strategies that maximize the diagnostic potential of this technology. PUBLIC HEALTH RELEVANCE: A sophisticated and widely used method for diagnosing and monitoring lung disease is computed tomography (CT scan). Usually, CT scans are evaluated in a subjective and qualitative manner, often resulting in variable assessment. The extensive digital information embedded within each scan is under-utilized and potential diagnostic information may remain unexplored. In addition, the lung is a unique non-solid, elastic organ that is highly variable in volume and sensitive to asymmetric distortion due to uneven disease involvement;architectural distortion can make it difficult to match the same lung regions on successive scans. There is a need for objective analytical strategies that fully exploit the information content of the CT scan to permit earlier detection and diagnosis of lung pathology. We have developed and tested a relatively simple semi-automated method for separating the lung into lobes, and analyzing local air and tissue contents within each lobe at a standard orientation and scale. We propose that this systematic analysis of CT scan provides objective indices that reflect clinical abnormalities in lung disease. We will quantify, map and compare CT-derived parameters within and among lobes in normal subjects and in patients with different severities of emphysema and interstitial lung disease. We will also correlate the CT-derived parameters with clinical symptoms and signs, and with lung function tests. These studies will establish or refute the usefulness of quantitative analysis in evaluating CT scan of the lung;they provide the essential data that may lead to future larger-scale studies to test whether this approach can predict disease progression or response to treatment. The extensive CT database maintained by LTRC provides a unique and valuable resource for addressing these issues. Given the widespread use (and possible overuse) of CT scan in pulmonary medicine and the small but real risk of harm associated with repeated radiation exposure, it is imperative to develop and adopt strategies that maximize the diagnostic potential of this powerful technology. (End of Abstract)